CSS3

From the October F2F CSS Working Group minutes (Apple’s proposal); Apple presented their proposals for Animations and Transitions, made remarks on Transforms, gradients, reflections. All four major browser vendors are interested in these proposals, and it is very likely that the CSSWG will accept to work on them. Exciting stuff!

Hakon presented a proposal for a border-parts property, which defines an on/off mask over the border as a series of lengths; you can check out his proposal in more detail on www-style. Note, this is still very much an unapproved proposal and the WG have stated that they are still unsure how usable/useful a solution such as this would be.

Dave Hyatt and Daniel Glazman initially came up with a proposal for CSS Variables, and fantasai has now created a counter-proposal detailing plans for a purely parse-time, non-scriptable solution. To help distinguish between the merits of both, I’ve come up with a brief article comparing and summarising both proposals.

An advantage of not posting for a while is that there are usually plenty of good subjects to talk about when you get back to it…

With a kind nod of the head to our post from way back in 2006, Andy Clarke has posted a screencast (with full transcript and code examples) of RGBa values in action on his website For A Beautiful Web.

Over at Javascript Kit they’ve produced an exhaustively in-depth look at structural pseudo-classes.

On the subject of Javascript, Eric Meyer has written about how it can be leveraged to extend CSS3 support across browsers.

And Helen from Helephant.com writes a good introduction to the box-sizing property.

Mozilla employee John Daggett has provided some try‐out builds of Firefox with support for the @font-face from CSS3’s web-fonts module. Currently available for Windows and Mac only—no Linux build yet—there remain several caveats as described in his comment on bug 70132, the most important of which being that the same‐site origin restriction is turned on by default, which means that most examples on the web will not work until you turn it off.

This information was released today on the Working Group blog and there looks to be some major updates.

• The shorthand syntax for border-radius has been updated so that you can now specify different irregular curves on an element – this was a well-known abiguity that Peter pointed out in the former Working Draft.
• Handy new background-position syntax for positioning offsets from corners other than the top left
• Spread values for box-shadow
• Box-shadows can now be inverted, to appear ‘inside’ an element thanks for the new inset keyword.
• Addition to background-color value syntax which means fallback colours can be specified
• Addition of no-clip value for background-clip
• Two additional background-size values (cover and contain; they both scale the image whislt retaining its intrinsic aspect ratio. The difference however, is that by using the cover keyword value, the image scales to the largest size such that both its width and its height can completely cover the background positioning area to. Whereas scale scales the image to the largest size such that both its width and its height can fit inside the background positioning area.
• New syntax for the background shorthand property

I’ll try and clarify some of these updates when I get some more time. All in all, some great new updates, though!

As you can’t have failed to notice, Google released their Chrome browser today. Chrome is based on the same version of Webkit as Safari 3.1 so should in theory have the same level of CSS support, although based on the very brief usage I’ve had of it so far it seems that text-shadow and @font-face aren’t working.

Update: I should add, of course, that this is still Beta software, and these issues may well be fixed before launch.

Anyone else noticed any missing features?

While browsing TechCrunch this morning, I stumbled across the site for iPhone developers tap tap tap, who had published their sales figures for iPhone apps in the first month. While these were very interesting, what’s even more interesting is the absolutely fabulous way in which these guys are using @font-face.

If you check their stylesheet, you’ll see this:

/* A font by Jos Buivenga (exljbris) -> www.exljbris.nl */
@font-face {
	  font-family: "Fertigo";
	  src: url(http://www.taptaptap.com/Fertigo.otf)
	  	  format("opentype");
}

Which renders into great looking text like this (at least in my Safari):

Kudos to the tap tap tap people, and if you see more of these in the wild, be sure to let us know!

Just a flying update, to provide some links of interest (with little-to-no comment):

• Firefox 3.1’s release date has been pushed back a little; I understand that, CSS-wise, Transforms and Web Fonts are the current blockers.
• John Resig takes a look at the implementation of border-image in FF 3.1 (with examples supplied by us)
• The CSS Marquee module has made it to Working Draft status. And now, with the rise of the mobile web and limited screen size, it actually has a practical implementation.

A first Alpha of Mozilla’s new browser is due for release shortly (probably tomorrow). Firefox 3 saw a lot of work go into speed, stability and the interface, but was slightly disappointing for front-end developers and saw CSS implementation overtaken by Safari and Opera.

3.1 will make up for that with a whole raft of features on their way. Implemented in the current nightlies (and therefore, most likely, in the Alpha) are:

• All remaining Selectors
• Text Shadow
• Box Shadow
• Border Image
• Column Rule

Planned for implementation but not yet landed features include:

• Media Queries
• Transforms

For the stat-hungry amongst us, 3.1 currently scores 84 on the Acid 3 test, and the implementation of Media Queries should take that up to 85.

There’s no indication of it, but I personally would like to see their border-radius syntax brought into line with the standard, and Transitions implemented to complement Transforms. I’ve been playing with Transitions recently, and they’re very cool.

What new CSS features would you like to see in Firefox 3.1?

The CSS3 Colour module is one of the most implemented CSS3 modules. This was previously in the Candidate Recommendation stage, but has just recently been reverted to Last Call. While this sounds like a step backwards, it was done due to the specification being updated.

The new version of the CSS3 Colour Module has removed those features that were not widely implemented, such as the flavor system colour and the @color-profile at-rule. These dropped features are now in a request for implementation, which basically means the W3C wants browser or user agent vendors to implement the features or they will not be included in the final recommendation. If this is the case, then they will either be dropped completely or moved to CSS Colour level 4. The last call lasts until the 1^st of September. If you’d like to give comments then send them to the www-style mailing list. The CSS Snapshot 2007 is waiting for this spec to go to Candidate Recommendation before it moves from Working Draft to Candidate Recommendation itself.

Now that the colour profile and flavor features have been removed, support is almost complete in three out of the four major browser engines. Firefox 3 supports the entire spec. Safari 3 has a bug with mixed values in RGB and RGBA, and does not support the the currentColor value, but the latest nightlies fixes both of these issues. Opera 9.5 does not support HSLA or RGBA, and doesn’t support the transparent value in CSS3 context (such as on the color property), but these are supported in the ACID3 build of Opera, and will be included in version of Opera that will use Core-2.2.

The beauty about working on a site that is specifically targeted for one the most popular mobile devices around (aka the iPhone) is that the vendor with the least CSS support (*cough* IE *cough*) doesn’t exist on it (purely because of its inadequate CSS support in this case).

Depending on the browser matrix you have to work with and the platforms you’re coding for, utilizing Level 3 (and some Level 2.1 in IE’s case) selectors simply isn’t realistic, although properties on the other hand can be used to progressively enhance an elements appearance in this particular context.

But what if a site was created to serve one specific handheld device in which Safari (along with its sufficient CSS3 support) was the default browser? Well, this is what Facebook has done with it’s iPhone-specifc UI. By peaking into their stylesheet you’ll notice that they’re implementing a number of Level 3 selectors, properties and property value additions. They include the:-

• User-select property (Webkit prefix)
• Negation Pseudo-class Selector
• RGBA color values (in conjunction with ‘background-color’ and ‘text-shadow’ properties
• Text Overflow property
• Box-sizing property (including Mozilla and Webkit prefixes)
• Border-image property (Webkit prefix)
• Text-shadow property
• Border-radius property
• Multiple Backgrounds
• Opacity

For comedy value, try viewing the site in IE.

Having this sort of free reign on selectors and properties is an interface developers dream in this day and age, and hopefully in a couple of years we’ll be able to structure a stylesheet with some similarly advanced features.